Mega ST Drive access = bombs

[SophieRose]

I've powered up my spare mega ST board and it's exhibiting a strange fault.

If I let it go to the desktop with no disk in I can use GEM but as soon as I stick a floppy in and try to access it I get a white screen and 4 bombs, same goes if I plug in the ultrasatan and try to boot of that... 4 bomb maybe more sometime but a white screen and bombs everytime for any kind of disk access.

So far I've swapped all the socketed chips / PSU / Floppy drive with my working mega board and it's still the same.

I changed

DMA
Shifter
Glue
MMU
CPU
WDC

Any pointers would be ace. I'm off to get schematics.

[exxos]

You need to start by doing all the mandatory fixes https://www.exxosforum.co.uk/forum/viewforum.php?f=17

[rubber_jonnie]

I'd agree with the mandatory fixes listed by exxos, the WD1772 data line pull up (Soldered directly to the WD1772, https://www.exxosforum.co.uk/forum/viewt ... 417#p15270) and the rest of the data line pull ups.

In my recent experience with a Mega and a non functional FDD, I found the signals on the WD1772 data lines to be very low, and the addition of the resistor pack to the WD1772 helped solve that.

In my case it was the DMA that was faulty in the end, so not likely to be your problem because you've swapped out the chip for a known working one.

There are some signals that get inverted on their way to the FDD, so when you trace the signals on the schematic, keep an eye out for this.

It's also worthwhile buzzing out the tracks from the WD1772 to their destinations as some were damaged on mine. It looked like the 1772 was replaced prior to my ownership, to try and fix the FDD problem and some of the vias and tracks were damaged, so bodge wiring was necessary.

If you don't have a diag cartridge I'd suggest one, since it has some useful and repeatable FDD tests, so you can watch what the motor and heads are doing in a consistent manner.

When I was tracing mine, I made up a spreadsheet with all the connections and where they went to, plus the pinout info for the WD1772, which I've attached, just remove the .txt extension. Hopefully it'll help.

[SophieRose]

Cheers for those pointers and the spreadsheet, should really help thank you. At least this will keep me busy during lockdown. :D

[rubber_jonnie]

Cheers for those pointers and the spreadsheet, should really help thank you. At least this will keep me busy during lockdown. :D

You're welcome.

I spent quite a lot of time troubleshooting the floppy problems on mine, and though I did have different symptoms and ultimately a dead DMA, a lot of the troubleshooting would be the same.

I'll keep an eye on this one in case I can help some more :)

[Darklord]

I had the same kind of problem with my Mega ST4 that runs my BBS once.

Anytime I tried to access the floppy drive, it would crash and bomb, sometimes rebooting.

Turned out to be the P/S in my case. Bought a new one and it all went away.

Hope you get it fixed soon! :)

[SophieRose]

Well my diag cart arrived, cheers Chris.

So it seems this mega board has two problems which may be related or not, time shall tell.

1. the blitter. If I install the blitter and un bridge the MB contacts as required, the blitter will lock up the blitter test and also the desktop when I move the mouse.

2. Floppy / ASCI use. bombs and a reeboot. I tried the diag cart and the floppy test failed and returned this. (see attached)

I shall delve deeper, if anyone has any ideas....

Soph

[exxos]

Did you do all the fixes ?

[SophieRose]

It's a good place to start it seems, I may as well do both my boards while I'm at it. Looks like more parts to order :)

[rubber_jonnie]

It is worth also buzzing out the tracks from point to point based on the schematics to make sure there are no damaged tracks.

Do you have access to an oscilloscope? If not a logic probe or multi-meter can be used to check logic levels are either 5V or 0V. I'd advise following the schematic and repeating the test and checking the output on each pin of the WD1772 to see if it matches the correct state based on the table I included with the spreadsheet I uploaded previously. Obviously this will take a while but it will verify operation. Also check that the state is changing at other connected points on the schematic, and be aware there are some level changes due to inverters.

Also make sure that the signals expected at the FDD do get there, and those that are generated by the floppy, such as write-protect, are seen at the WD1772.

It's unfortunately a painstaking process, and I would also advise as per @exxos to do the fixes first, particularly the bus pull up resistor packs and adding a pull up to the WD1772.

On my Mega 4, one of the bus pull up resistor packs had a crack down the middle that wasn't visible until I removed it, so wasn't working as a pull up for half the bus, so it is worth replacing them.